From fd2fd60f7c88ce8b50a5d491fd9b1b7c5bc960d8 Mon Sep 17 00:00:00 2001
From: Strahinja Val Markovic <val@markovic.io>
Date: Thu, 21 Feb 2013 22:12:34 -0800
Subject: [PATCH] Identifier collection now smarter about strings

Previously, a string like 'foo\'bar\'zoo' would make the collection process
think that "bar" is not inside a string because it wouldn't recognize that the
quotes are escaped. Now it does.

Fixes #143.
---
 cpp/ycm/IdentifierUtils.cpp            |  6 ++++--
 cpp/ycm/tests/IdentifierUtils_test.cpp | 18 ++++++++++++++++++
 2 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/cpp/ycm/IdentifierUtils.cpp b/cpp/ycm/IdentifierUtils.cpp
index fbfac6cb..b559b9be 100644
--- a/cpp/ycm/IdentifierUtils.cpp
+++ b/cpp/ycm/IdentifierUtils.cpp
@@ -30,9 +30,11 @@ const char *COMMENT_AND_STRING_REGEX =
   "|"
   "/\\*.*?\\*/"  // C-style comments, '/* ... */'
   "|"
-  "'[^']*'" // Anything inside single quotes, '...'
+  "'(?:\\\\'|.)*?'" // Anything inside single quotes, '...', but mind the
+                    // escaped quote
   "|"
-  "\"[^\"]*\""; // Anything inside double quotes, "..."
+  "\"(?:\\\\\"|.)*?\""; // Anything inside double quotes, "...", but mind
+                        // the escaped double quote
 
 const char *IDENTIFIER_REGEX = "[_a-zA-Z]\\w*";
 
diff --git a/cpp/ycm/tests/IdentifierUtils_test.cpp b/cpp/ycm/tests/IdentifierUtils_test.cpp
index 04f16a57..48092c1e 100644
--- a/cpp/ycm/tests/IdentifierUtils_test.cpp
+++ b/cpp/ycm/tests/IdentifierUtils_test.cpp
@@ -63,6 +63,15 @@ TEST( IdentifierUtilsTest, RemoveIdentifierFreeTextWorks ) {
                 "bar \n"
                 "qux" );
 
+  EXPECT_STREQ( RemoveIdentifierFreeText(
+                  "foo \n"
+                  "bar 'fo\\'oz\\nfoo'\n"
+                  "qux"
+                ).c_str(),
+                "foo \n"
+                "bar \n"
+                "qux" );
+
   EXPECT_STREQ( RemoveIdentifierFreeText(
                   "foo \n"
                   "bar \"foo\"\n"
@@ -71,6 +80,15 @@ TEST( IdentifierUtilsTest, RemoveIdentifierFreeTextWorks ) {
                 "foo \n"
                 "bar \n"
                 "qux" );
+
+  EXPECT_STREQ( RemoveIdentifierFreeText(
+                  "foo \n"
+                  "bar \"fo\\\"oz\\nfoo\"\n"
+                  "qux"
+                ).c_str(),
+                "foo \n"
+                "bar \n"
+                "qux" );
 }